Banca de DEFESA: FRANCISCO GUARACI GOMES DE OLIVEIRA

Uma banca de DEFESA de MESTRADO foi cadastrada pelo programa.
STUDENT : FRANCISCO GUARACI GOMES DE OLIVEIRA
DATE: 19/06/2024
TIME: 14:00
LOCAL: Remoto - videoconferência
TITLE:

Erosion and Reforestation in the Caatinga: Application of RUSLE3D and USPED Combined with Artificial Intelligence

KEY WORDS:

Seridó, Tropical Dry Forest, Sustainability, Genetic Algorithm

PAGES: 87
BIG AREA: Ciências Agrárias
AREA: Recursos Florestais e Engenharia Florestal
SUBÁREA: Conservação da Natureza
SPECIALTY: Conservação de Bacias Hidrográficas
SUMMARY:

The present study addresses the issue of soil erosion within the Caatinga biome, emphasizing how land use and cover changes, along with inadequate management practices, have accelerated environmental degradation. The expansion of agricultural and livestock activities and the reduction of native vegetation have increased soil exposure to erosion, potentially compromising the productivity and sustainability of local ecosystems. Furthermore, soil degradation contributes to biodiversity loss and the release of carbon into the atmosphere, thereby exacerbating climate change. To confront these challenges, sustainable management strategies and the application of advanced technologies, such as predictive models and artificial intelligence, are imperative for identifying priority areas for intervention and optimizing conservation efforts.The objective of this study was to analyze the impact of land use and cover changes on erosive processes in the Caatinga and to assess the effectiveness of reforestation strategies, developed using artificial intelligence, in controlling erosion. In the first article, the RUSLE3D model was applied to the Seridó River Basin (BHRS) to analyze soil erosion between the years 1988 and 2018. The study revealed a significant reduction in native vegetation, with a 15.12% decrease in Caatinga Forest. Additionally, more than 35% of the studied area underwent changes in land use, primarily due to the expansion of agricultural activities. Despite an overall decrease in average soil loss over time, a correlation between the reduction in vegetation cover and increased erosion was noted. Hence, implementing sustainable management practices is crucial for controlling erosion in the Caatinga, with a focus on slopes and exposed soil areas, which are the most vulnerable.The second article proposed forest restoration in the Caatinga using optimization algorithms to identify priority areas for reforestation. The application of the USPED model and genetic algorithms, considering a proposed 20% increase in forest cover, led to a reduction in soil loss from 105 thousand to 75 thousand tons per year. The reforestation proposed by the algorithm enhanced soil stability, increasing stable areas from 39.48% to 43.07%, but resulted in increased fragmentation, which could negatively impact carbon storage and exacerbate edge effects. Thus, while artificial intelligence proved effective in optimizing forest restoration, it is essential to balance erosion mitigation with the maintenance of connectivity between vegetation fragments. The findings of this study are expected to provide valuable insights for improving environmental sustainability and health within the Caatinga biome of the BHRS.

COMMITTEE MEMBERS:
Presidente - 1305184 - JHONES DA SILVA AMORIM
Externo à Instituição - DANILO PAULUCIO DA SILVA - UESB
Externo à Instituição - RUBENS JUNQUEIRA - UFLA